One-half of all pregnancies concluding in 1994 were unintended and one-half of these pregnancies were terminated by abortion. Opinion polls on abortion do not seem to reflect the behavior of 1.5 million women who have an abortion each year. We need to distinguish the generalized attitudes in the public from the personal attitudes held by women. The demand for abortion in the U.S. is likely to remain high. Therefore, it is important to understand the formation of and the change in abortion attitudes because attitudes shape intentions and intentions determine behavior. Equally important is the males' attitudes toward abortion because: First, policy making in U.S. is dominated by men. Second, men are equally responsible for unintended pregnancies, and their attitudes toward abortion may influence their own, and their partners' sexual and contraceptive behavior. Third, men are also affected by unintended pregnancy and they can influence decisions about pregnancy resolution. The primary purpose of the proposed research is to analyze the cohort, gender, and race differences in abortion attitudes, in the factors influencing the formation of abortion attitudes, and in the effects of abortion attitudes on subsequent reproductive behaviors. Specifically, we will examine: (1) Cohort differences in abortion attitudes among women who were 20-27 years of age in 1983 and in 1991; (2) The differences in the antecedents of abortion attitudes for the 1983 and the 1991 cohorts of women focusing on the factors that affect attitude formation; (3) Race differences in the abortion attitudes and in the antecedents of the abortion attitudes among women in 1983 and in 1991; (4) The change in abortion attitudes of the 1983 cohort as the cohort ages with particular focus on the effects of life events during the observation period; (5) Gender differences in the abortion attitudes and the antecedents of abortion attitudes in 1991; (6) Race differences in the abortion attitudes and in the antecedents of the abortion attitudes among men in 1991; (7) The relationship between abortion attitudes and subsequent reproductive behaviors (e.g., contraceptive choice, unintended birth, abortion); and (8) Gender, race and cohort differences in such behaviors as attributable to the differences in abortion attitudes. The investigators will use data from the 1983 National Survey of Unmarried Women (NSUW), the 1991 National Survey of Women (NSW), the 1991 National Survey of Men (NSM), the 1993 follow-up found of the NSM, and the 1993 and 1995 follow-up rounds of the NSW to accomplish these aims. Confirmatory factor analysis will be used to obtain the measurement model. Regression and structural equations will be used to estimate the conceptual model.